Hybrid Regression Research Articles

Grandparental childcare and subjective well-being: The role of activities and reasons for care.

Despite extensive research on the impact of grandchild care on grandparents' well-being, no studies have examined the frequency of activities that grandparents engage in with their grandchildren or the reasons for care using nationally representative data. We address this gap using waves 8 (2016/2017) and 9 (2018/19) of the English Longitudinal Study of Ageing, a nationally representative study of English older adults. We employ hybrid regression models to distinguish between within- and between-individual effects on grandparents' subjective wellbeing, as captured by quality of life and depressive symptoms. Our findings show that the mostly positive relationships between frequency of grandparental activities and reasons for care and well-being are largely due to between-grandparent differences rather than within-grandparent changes over time. Within-grandparent analyses show a mostly non-significant impact of increased grandparental activities on well-being, except for a detrimental effect when grandchildren stay overnight or are cared for when sick. The influence of reasons for care are more nuanced. Changes in reasons, such as giving parents a break or allowing them to go out in the evening, are significantly associated with a worsening of elevated depressive symptoms. Conversely, changes such as helping grandchildren develop as people or feeling engaged with young people improve the quality of life for the same grandparents over time. When grandparental care is provided because it is difficult to refuse we find a negative association with well-being between grandparents. Overall, our findings highlight the importance of more detailed information on grandparent-grandchild interactions and reasons for care to better understand well-being outcomes.

Determinants of traditional Korean medicine utilization by individuals with physical disabilities: inference of causal relationships using Korea health panel data

BackgroundThe demand for health management services has grown among individuals with physical disabilities. It is noteworthy that a significant proportion of this demographic has sought the services of traditional Korean medicine (TKM). Nevertheless, there is a lack of research on the characteristics of TKM utilization within this population. Therefore, this study aims to explore the characteristics of individuals with physical disabilities who utilize TKM services, to support future policy development in TKM for enhancing the health and quality of life for individuals with physical disabilities.MethodsThe study utilized a dataset from the Korea Health Panel (KHP) version 1.7, from the years 2008 to 2018. A panel analysis was conducted to analyze the correlation between the characteristics of participants and their influence of the TKM utilization.ResultsBased on the panel logistic regression analysis, the probability of TKM utilization was higher for individuals with physical disabilities who were female (p < 0.001), had a chronic musculoskeletal disorder (p < 0.001), Charlson Comorbidity Index of 1 (p = 0.025), and multiple chronic conditions (p = 0.025). Based on the panel hybrid regression analysis, those covered with Medical Aid spent less on TKM than those covered with National Health Insurance (p < 0.001), while those who were female (p < 0.001), had chronic musculoskeletal disorders (p = 0.007), or multiple chronic conditions (p = 0.007) had higher medical expenses for TKM utilization.ConclusionsIndividuals with physical disabilities accompanying nonfatal chronic musculoskeletal disorders or other chronic conditions frequently utilized TKM services. Those covered by Medical Aid spent less on TKM, while females and individuals with chronic musculoskeletal disorders or multiple chronic conditions incurred higher medical expenses for TKM. These results suggest the need for targeted interventions and policy changes to improve TKM access and address financial burdens for people with disabilities.

Monitoring Cover Crop Biomass in Southern Brazil Using Combined PlanetScope and Sentinel-1 SAR Data

Precision agriculture integrates multiple sensors and data types to support farmers with informed decision-making tools throughout crop cycles. This study evaluated Aboveground Biomass (AGB) estimates of Rye using attributes derived from PlanetScope (PS) optical, Sentinel-1 Synthetic Aperture Radar (SAR), and hybrid (optical plus SAR) datasets. Optical attributes encompassed surface reflectance from PS’s blue, green, red, and near-infrared (NIR) bands, alongside the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI). Sentinel-1 SAR attributes included the C-band Synthetic Aperture Radar Ground Range Detected, VV and HH polarizations, and both Ratio and Polarization (Pol) indices. Ground reference AGB data for Rye (Secale cereal L.) were collected from 50 samples and four dates at a farm located in southern Brazil, aligning with image acquisition dates. Multiple linear regression models were trained and validated. AGB was estimated based on individual (optical PS or Sentinel-1 SAR) and combined datasets (optical plus SAR). This process was repeated 100 times, and variable importance was extracted. Results revealed improved Rye AGB estimates with integrated optical and SAR data. Optical vegetation indices displayed higher correlation coefficients (r) for AGB estimation (r = +0.67 for both EVI and NDVI) compared to SAR attributes like VV, Ratio, and polarization (r ranging from −0.52 to −0.58). However, the hybrid regression model enhanced AGB estimation (R2 = 0.62, p &lt; 0.01), reducing RMSE to 579 kg·ha−1. Using only optical or SAR data yielded R2 values of 0.51 and 0.42, respectively (p &lt; 0.01). In the hybrid model, the most important predictors were VV, NIR, blue, and EVI. Spatial distribution analysis of predicted Rye AGB unveiled agricultural zones associated with varying biomass throughout the cover crop development. Our findings underscored the complementarity of optical with SAR data to enhance AGB estimates of cover crops, offering valuable insights for agricultural zoning to support soil and cash crop management.

Hybrid regression analysis for the static strength of seasonally frozen soils

Hybrid regression analysis for the static strength of seasonally frozen soils

Prediction of California bearing ratio using hybrid regression models

Prediction of California bearing ratio using hybrid regression models

Visualization Study of Oil Well Fracturing Production Prediction Model Based on Weighted Hybrid Regression Algorithm

This study aims to explore visualization methods for an oil well fracturing production prediction model based on a weighted hybrid regression algorithm. The objective is to construct a mathematical model capable of accurately predicting oil well fracturing production through theoretical analysis and to enhance the model's interpretability and applicability using visualization techniques. The research primarily employs a weighted hybrid regression algorithm, which combines the advantages of multiple regression and weighted least squares to effectively address heteroscedasticity and multicollinearity issues in the data. Initially, a systematic analysis of factors influencing oil well fracturing production, such as geological conditions, fracturing parameters, and well structure, is conducted. Based on these factors, a weighted hybrid regression model is constructed. Subsequently, mathematical derivation and simulation experiments are used to validate the model's predictive accuracy and stability. Finally, visualization techniques, such as 3D graphics and dynamic simulation, are utilized to present the model’s prediction results and trends, enhancing the model’s intuitiveness and comprehensibility. The study concludes that the weighted hybrid regression algorithm demonstrates high accuracy and robustness in predicting oil well fracturing production, and visualization techniques effectively assist decision-makers and engineers in understanding and managing the oil well production process.

Rent Price Prediction with Advanced Machine Learning Methods: A Comparison of California and Texas

The forecast of rent prices in dynamic housing markets is of fundamental importance to renters, landlords, investors, and politicians alike. Machine learning models offer flexibility, excel at modeling complex relationships, and provide outstanding forecast precision. This study compares advanced machine learning models, extreme gradient boosting regressor (XGBoost), light gradient boosting machine (LightGBM), random forest, ridge regression, and two ensemble approaches, to predict California and Texas rent prices. The two ensemble approaches include a hybrid regression of averaging base models and a 2-level stacked generalization model. The results revealed that a stacked generalization ensemble with base models random forest, XGBoost, LightGBM, and meta-model ridge regression achieved the best performance for the California dataset by generating the lowest MSE and highest R2 of 46116.3 and 0.8858, respectively. In contrast, random forest outperformed both ensemble models with the lowest MSE and MAPE of 18401.93 and 9.7003%, respectively, and the highest R2 of 0.7992. These methodologies can assess future rental property worth, serve as indicators for market dynamics, and aid in establishing real estate policies, thereby providing practical guidance to individuals, businesses, and policymakers.

Preference-based utility weights for the Individualized Neuromuscular Quality of Life Questionnaire (INQoL), with a focus on non-dystrophic myotonia (NDM)

IntroductionThe Individualized Neuromuscular Quality of Life Questionnaire (INQoL) is used to measure quality of life in neuromuscular disorders such as non-dystrophic myotonia (NDM). Here we report methods to estimate utilities, with a focus on NDM, from this questionnaire based on two preference elicitation exercises.MethodsEight items from the INQoL were selected with input from three neuromuscular disorder clinical experts with expertise in treating NDM. A discrete choice experiment (DCE) survey of UK general public respondents (n = 508) described outcomes defined by the INQoL items. The same 8 items were also valued using time trade-off (TTO) face-to-face interviews (n = 200). A hybrid regression modelling approach combined both datasets to inform the utility weights.ResultsHybrid modelling of DCE and TTO data in conjunction improved out-of-sample predictive accuracy. The selected INQoL utility model indicates substantial disutility associated with all eight dimensions of health, with the greatest losses associated with subjective items such as pain and depression.DiscussionThe hybrid modelling approach allows us to combine data from the two methodologies and maximize the information from each to inform the utility weights for the INQoL. The TTO is the more conventional valuation method, but combined with the larger DCE study produced better descriptive coverage. This is a relatively novel method for estimating weights which we think is particularly well suited to economic evaluations of orphan drugs.

Comprehensive hybrid regression model for financial forecasting in neutrosophic logic

Regression analysis is a widely used tool in several fields. In this paper, we propose a comprehensive, multistep regression model for financial forecasting. The proposed hybrid model combines preprocessing, feature selection, and cross-validation to obtain a powerful approach to forecasting. The extension of the proposed model to neutrosophic sets is discussed. The model is applied to the case study of real estate prices. The results demonstrate the efficacy of the model.

Effluent parameters prediction of a biological nutrient removal (BNR) process using different machine learning methods: A case study

This paper proposes a novel targeted blend of machine learning (ML) based approaches for controlling wastewater treatment plant (WWTP) operation by predicting distributions of key effluent parameters of a biological nutrient removal (BNR) process. Two years of data were collected from Plajyolu wastewater treatment plant in Kocaeli, Türkiye and the effluent parameters were predicted using six machine learning algorithms to compare their performances. Based on mean absolute percentage error (MAPE) metric only, support vector regression machine (SVRM) with linear kernel method showed a good agreement for COD and BOD5, with the MAPE values of about 9% and 0.9%, respectively. Random Forest (RF) and EXtreme Gradient Boosting (XGBoost) regression were found to be the best algorithms for TN and TP effluent parameters, with the MAPE values of about 34% and 27%, respectively. Further, when the results were evaluated together according to all the performance metrics, RF, SVRM (with both linear kernel and RBF kernel), and Hybrid Regression algorithms generally made more successful predictions than Light GBM and XGBoost algorithms for all the parameters. Through this case study we demonstrated selective application of ML algorithms can be used to predict different effluent parameters more effectively. Wider implementation of this approach can potentially reduce the resource demands for active monitoring the environmental performance of WWTPs.

A relationship between vehicle coastdown performance metrics and fuel efficiency

In contemporary times, societal and legal institutions are earnestly committed to mitigating carbon dioxide (CO2) emissions into the atmosphere, thereby averting global warming. The automotive industry mirrors this commitment, necessitating the development of vehicles characterized by heightened energy efficiency. Such vehicles should boast an extended range while minimizing fuel consumption and CO2 emissions. A pivotal aspect in vehicle development involves the early determination of fuel economy (FE), contingent upon the projected characteristics of the vehicle and its intended market positioning with regard to fuel economy. Over time, diverse methodologies for calculating FE have emerged. These include hybrid regression models incorporating key input variables such as instantaneous vehicle speed and acceleration measurements, theoretical approximation methods derived from the physical properties of engine-vehicle systems, correlations among traffic-related parameters, power-demand models, as well as artificial neural networks and genetic algorithms. These approaches leverage various vehicle input data, encompassing engine speed, torque, fuel flow, intake manifold mean temperature, make of the car, engine style, weight of the car, vehicle type, transmission system type, and other relevant factors. Regrettably, the literature lacks a straightforward and expeditious means of estimating FE for a group of analogous vehicles, a deficiency addressed by the present work. The proposed empirical physical model facilitates FE estimations for a vehicle using a parsimonious set of input information: engine displacement, performance (recovery speed times), and road load data. These inputs are subjected to mathematical parameter adjustments within the empirical model. To validate the efficacy of this approach, the empirical model is applied in a Brazilian case study. The results attest to the success of the method, demonstrating a margin of error within 6% when compared to the official Brazilian government public data for the tested vehicle, particularly in the contexts of city and highway cycles.

Income distribution and nudity on social media: Attention economics of Instagram stars

AbstractSocial media stars gain star‐status with uploads on social media pages like YouTube, TikTok, or Instagram. One of the most popular platforms is “Instagram” owned by Meta/Facebook. The growing social, cultural, and economic power of so‐called influencers raises questions about key drivers of success and, moreover, distribution of income on social media platforms. Instagram has been accused of strategically favoring images with nude content. In order to shed light on this socio‐critical aspect, this paper examines the following research questions: Does body exposure drive income success on Instagram? Is there a difference between male and female content in this regard? This paper empirically analyzes 500 top Instagram stars within the categories (1) fashion and beauty, (2) fitness and sports, (3) music, (4) photo and arts, and (5) food and vegan. The data provide information on popularity, posting behavior, and price estimates per post. Using hybrid regression models, the results show indeed positive impact of body exposure on monetary success. Accounts with high level of body exposure achieve higher prices and advertising revenues than accounts with less nudity, regardless of the gender. Regarding gender differences, male content achieves on average higher advertising prices, whereas female accounts provide more branded content and eventually achieve higher advertising revenues.

Hybrid regression models: predicting of the capillary water absorption properties of construction stones

Hybrid regression models: predicting of the capillary water absorption properties of construction stones

Becoming Sandwiched in Later Life: Consequences for Individuals' Well-Being and Variation Across Welfare Regimes.

The experience of being sandwiched between support obligations towards both aging parents and adult offspring is likely to become more common and more relevant. We aim at assessing the effect of demographic and social sandwiching on the psychological health and subjective well-being of individuals experiencing these transitions, and to what extent, these effects vary across welfare regimes. Data are from 63,585 individuals aged 50-75 participating in the Survey of Health, Ageing and Retirement in Europe (SHARE). We estimate within- and between-individual effects using hybrid regression models to predict depressive symptoms (EURO-D) and subjective well-being (control, autonomy, self-realization, and pleasure [CASP]). Among demographically sandwiched women, transitioning into social sandwiching and into supporting only parents was associated with a moderate but statistically significant increase in EURO-D and decline in CASP scores. The same association is not observed for male respondents. The pattern of variation among women living in countries characterized by different welfare regimes suggests that social sandwiching is less detrimental in Nordic regimes than in other welfare contexts. Results from the between-individuals part of the model indicate that there is a selection into social sandwiching of more healthy individuals into support roles. However, the within-individuals part of the model indicates that the transition into social sandwiching has a detrimental effect on women's (but not men's) psychological health and well-being. The explanations for this gendered effect of social sandwiching may be found in the "invisible" support provided by women and the gendered division of specific care tasks.

하이브리드 회귀모델을 이용한 태양광발전이 통합된 배전선로의 순부하 예측 방법

하이브리드 회귀모델을 이용한 태양광발전이 통합된 배전선로의 순부하 예측 방법

The EQ-5D-3L Valuation Study in Pakistan.

To utilize EQ-5D in economic evaluations, a societal-based value set is needed. To date, no value sets exist for any EQ-5D instrument in Pakistan. Previous EQ-5D studies conducted in Pakistan 'borrowed' health preferences developed in other countries. However, for a value set to be valid for Pakistani population, it should represent the preferences of the Pakistani population, and culture and living standards of Pakistan. The aim of this study was to derive a Pakistani EQ-5D-3L value set. A moderately representative sample aged 18 years and over was recruited from the Pakistani general population. A multi-stage stratified quota method with respect to ethnicity, gender, age and religion was utilized. Two elicitation techniques, the composite time trade-off (cTTO) and discrete choice experiments (DCE) were applied. Interviews were undertaken by trained interviewers using computer-assisted face-to-face interviews with the EuroQol Portable Valuation Technology (EQ-PVT) platform. To estimate the value set, a hybrid regression model combining cTTO and DCE data was used. A total of 289 respondents who completed the interviews were included for the analysis. The hybrid model correcting for heteroskedasticity without a constant was selected as the final model for the value set. It is shown that being unable to do usual activities (level 3) was assigned the largest weight, followed by mobility level 3, self-care level 3, pain/discomfort level 3 and anxiety/depression level 3. The worst health state was assigned the value - 0.171 in the final model. A Pakistani country-specific EQ-5D-3L value set is now available. The availability of this value set may help promote and facilitate health economic evaluations and health-related quality-of-life (HRQoL) research in Pakistan.

Estimating the monetary value of a Quality-Adjusted Life-Year in Quebec.

The value of a Quality-Adjusted Life-Year (QALY) is of great importance for the healthcare system. It helps when it comes to defining a cost-effectiveness threshold for the evaluation of health technologies. No willingness-to-pay value for a QALY exists in the province of Quebec, Canada. In this paper, we empirically investigated the monetary value of a QALY for the population of Quebec. Based on the Short-Form 6-Dimension version 2 (SF-6Dv2), we conducted an online survey with a representative adult sample living in Quebec. We used a time trade-off (TTO) combined with contingent valuation (CV), and a discrete choice experiment (DCE) to assess both the population's willingness to pay (WTP) for one QALY and the marginal WTP for health attributes. A health utility algorithm using hybrid regression was developed to determine a preference-based value set for health states. Main analysis was conducted on 993 answers for the CV and 2143 answers for the DCE. The willingness-to-pay per QALY varied from CA$ 47,048.84 (CI: 21,554.38; 72,543.30) for CV to CA$ 73,936.87 (CI: 63,105.40; 84,768.35) for DCE. Among the 6 dimensions of the SF-6Dv2, marginal WTP varied from CA$ 4499.15 (CI: 2975.06; 6023.25) for more role accomplishment in daily activities to CA$ 15,867.12 (CI: 13,825.75; 17,908.49) for less pain. Robustness check with multiple alternative samples, as well as alternative health utility algorithms, showed that the results were robust and the DCE method provided 50% larger results than the CV method, although confidence intervals overlap. This paper provides useful information for decision-makers on the monetary value of a QALY in Quebec.

Hybrid regression analysis for simulating the rheological and mechanical properties of SCC

Hybrid regression analysis for simulating the rheological and mechanical properties of SCC

Prediction of daytime and nighttime ground-level ozone using the hybrid regression models

Ozone is one of the major challenges for the air quality community due to its adverse impact on the environment and human health. This study seeks to improve the understanding of underlying mechanisms for several developed models for ozone prediction. We aim to establish a robust prediction model for ozone concentration up to the next four hours. Three years dataset including ozone (O3), nitrogen oxide (NOx), nitric oxide (NO), sulphur dioxide (SO2), nitrogen dioxide (NO2), carbon monoxide (CO), particulate matter (PM10, PM2.5), wind speed, solar radiation, temperature, and relative humidity (RH) were used in this study. The data were analyzed by using Multiple Linear Regression (MLR), Principal Component Regression (PCR), and Cluster-Multiple Linear Regression (CMLR) in predicting the next hours of O3 concentration. Results show that the MLR models executed high accuracy for O3t+1 (R2= 0.313), O3,t+2 (R2= 0.265), O3,t+3 (R2= 0.227) and O3,t+4 (R2= 0.217) as the best fitted-model. In conclusion, the MLR model is suitable for the next hour's O3 concentration prediction.

Hybrid ARDL-MIDAS-Transformer time-series regressions for multi-topic crypto market sentiment driven by price and technology factors

This paper develops a novel hybrid Autoregressive Distributed Lag Mixed Data Sampling (ARDL-MIDAS) model that integrates both deep neural network multi-head attention Transformer mechanisms, and a number of covariates, including sophisticated stochastic text time-series features, into a mixed-frequency time-series regression model with long memory structure. In doing so, we demonstrate how the resulting class of ARDL-MIDAS-Transformer models allows one to maintain the interpretability of the time-series models whilst exploiting the deep neural network attention architectures. The latter may be used for higher-order interaction analysis, or, as in our use case, for design of Instrumental Variables to reduce bias in the estimation of the infinite lag ARDL-MIDAS model. Our approach produces an accurate, interpretable forecasting framework that allows one to forecast end-of-day sentiment intra-daily, with readily attainable time-series regressors. In this regard, we conduct a statistical time-series analysis on mixed data frequencies to discover and study the relationships between sentiment from our custom stochastic text time-series sentiment framework, alternative popular sentiment extraction frameworks (BERT and VADER), and technology factors, as well as to investigate the role that price discovery has on retail cryptocurrency investors’ sentiment (crypto sentiment). This is an interesting time-series modelling challenge as it involves working with time-series regression models in which the time-series response process, and the regression time-series covariates, are observed at different time scales. Specifically, a detailed real-data study is conducted where we explore the relationship between daily crypto market sentiment (of positive, negative and neutral polarity) and the intra-daily (hourly) price log-return dynamics of crypto markets. The sentiment indices constructed for a variety of “topics” and news sources are produced as a collection of time-series capturing the daily sentiment polarity signals for each “topic”, namely each particular market or crypto asset. Different sentiment methods are developed in a time-series context, and utilised in the proposed hybrid regression framework. Furthermore, technology factors are introduced to capture network effects, such as the hash rate which is an important aspect of the money supply relating to the mining of new crypto assets, and block hashing for transaction verification. Throughout our real data study, we provide guidance and insights on how to use our hybrid model to combine—in a transparent, non-black-box way—covariates obtained with different time resolutions, how to understand the arising dynamics between these covariates, potentially under the presence of long memory structure, and, finally, successfully leverage these in forecasting applications. The hybrid model developed demonstrated superior performance to alternatives in both in-sample and forecasting application on real data.